DESCRIPTION (adapted from the Abstract): The purpose of this project is to link the theory and methods of epidemiological network modeling to empirical analyses of sexual networks and transmission of infection with the human immunodeficiency virus (HIV). Network modeling is used to study the impact of non-random mixing on the spread of disease. People employ complex rules for choosing sexual partners--selecting on attributes such as age, race, gender, and sexual preference, as well as a host of other socioeconomic signals--and the biases that structure their networks also operate to channel HIV. In general, the more extended and intimate the contact between partners, the stronger the selective bias in choosing them. Diseases that require intimate contact for transmission are thus affected by stronger network biases than diseases passed by casual contact. Network structure is now recognized to have an important role in the sexual spread of HIV and other sexually transmitted diseases (STD's). Although the pace of theoretical and methodological research in this area has been impressive, it has not been matched by empirical studies of network-mediated HIV transmission in different populations, or guided by the goal of finding effective and efficient interventions. These are the two goals of this project. The proposed project will involve detailed comparative analysis of two recent sexual network studies, one from Thailand (field period: 1992-3) and the other from Uganda (field period: 1994). The two countries represent different phases of the epidemic: young and mature, respectively. Both studies have highly comparable quantitative and qualitative data. The proposed analysis has five specific aims: (1) to refine existing network methods for epidemiological modeling; (2) to identify the key components of network structure for predicting HIV transmission dynamics; (3) to investigate the interaction between biomedical and network-related behavioral aspects of transmission; (4) to investigate the impact of network-channeled disease on demographic processes of fertility and mortality; and (5) to use the network-based projection methods to identify strategic points of intervention. The comparative approach adopted here should help to distinguish both the common and the culturally specific features that are important for understanding and intervention. Some aspects of the network transmission dynamics are expected to have fairly universal applicability (e.g., the importance of age-matching); comparative analysis will highlight how these aspects operate in specific contexts. Other aspects will be culturally specific (e.g., the organization of commercial sex, the patterns of simultaneous partnerships). Here, comparative analysis will identify the effects of such culturally-specific structures on the spread of the epidemic, and highlight the context sensitivity of intervention development. The goal is to take "network modeling" beyond its current status as a metaphor standing for an ad hoc collection of incompatible methods, and make it a more useful tool for understanding and intervention in epidemiology.